deliverable/linux.git
10 years agoASoC: Intel: don't select RT5640 if !I2C
Stephen Warren [Mon, 10 Mar 2014 16:38:40 +0000 (10:38 -0600)] 
ASoC: Intel: don't select RT5640 if !I2C

The rt5640 driver won't compile without I2C enabled. Hence, the Intel
Haswell and Baytrail+RT5640 ASoC drivers must also depend on I2C, since
these select RT5640.

This solves:
sound/soc/codecs/rt5640.c:2220:1: warning: data definition has no type or storage class [enabled by default]
sound/soc/codecs/rt5640.c:2220:1: error: type defaults to ‘int’ in declaration of ‘module_i2c_driver’ [-Werror=implicit-int]
sound/soc/codecs/rt5640.c:2220:1: warning: parameter names (without types) in function declaration [enabled by default]
sound/soc/codecs/rt5640.c:2210:26: warning: ‘rt5640_i2c_driver’ defined but not used [-Wunused-variable]

Reported-by: Jim Davis <jim.epost@gmail.com>
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
10 years agoASoC: pxa: tosa: Convert to table based DAPM and control setup
Lars-Peter Clausen [Wed, 12 Mar 2014 14:27:39 +0000 (15:27 +0100)] 
ASoC: pxa: tosa: Convert to table based DAPM and control setup

Use table based setup to register the controls and DAPM widgets and routes. This
on one hand makes the code a bit shorter and cleaner and on the other hand the
board level DAPM elements get registered in the card's DAPM context rather than
in the CODEC's DAPM context.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@linaro.org>
10 years agoASoC: pxa: magician: Convert to table based DAPM and control setup
Lars-Peter Clausen [Wed, 12 Mar 2014 14:27:38 +0000 (15:27 +0100)] 
ASoC: pxa: magician: Convert to table based DAPM and control setup

Use table based setup to register the controls and DAPM widgets and routes. This
on one hand makes the code a bit shorter and cleaner and on the other hand the
board level DAPM elements get registered in the card's DAPM context rather than
in the CODEC's DAPM context.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@linaro.org>
10 years agoMerge branch 'topic/dapm' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie...
Mark Brown [Thu, 13 Mar 2014 09:40:56 +0000 (09:40 +0000)] 
Merge branch 'topic/dapm' of git://git./linux/kernel/git/broonie/sound into asoc-pxa

10 years agoASoC: snappercl15: Convert to table based DAPM setup
Lars-Peter Clausen [Wed, 12 Mar 2014 14:27:36 +0000 (15:27 +0100)] 
ASoC: snappercl15: Convert to table based DAPM setup

Use table based setup to register the DAPM widgets and routes.  This on one hand
makes the code a bit shorter and cleaner and on the other hand the board level
DAPM elements get registered in the card's DAPM context rather than in the
CODEC's DAPM context.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@linaro.org>
10 years agoASoC: ams-delta: Fix compile error
Lars-Peter Clausen [Wed, 12 Mar 2014 14:27:29 +0000 (15:27 +0100)] 
ASoC: ams-delta: Fix compile error

snd_soc_dapm_mutex_unlock() wants a pointer to the DAPM context, not the CODEC.

Fixes: 03510ca07 ("ASoC: ams-delta: Update locking around use of DAPM pin API")
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@linaro.org>
10 years agoMerge tag 'asoc-v3.15' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound...
Takashi Iwai [Thu, 13 Mar 2014 08:53:25 +0000 (09:53 +0100)] 
Merge tag 'asoc-v3.15' of git://git./linux/kernel/git/broonie/sound into for-next

ASoC: Updates for v3.15

Quite a busy release for ASoC this time, more on janitorial work than
exciting new features but welcome nontheless:

 - Lots of cleanups from Takashi for enumerations; the original API for
   these was error prone so he's refactored lots of code to use more
   modern APIs which avoid issues.
 - Elimination of the ASoC level wrappers for I2C and SPI moving us
   closer to converting to regmap completely and avoiding some
   randconfig hassle.
 - Provide both manually and transparently locked DAPM APIs rather than
   a mix of the two fixing some concurrency issues.
 - Start converting CODEC drivers to use separate bus interface drivers
   rather than having them all in one file helping avoid dependency
   issues.
 - DPCM support for Intel Haswell and Bay Trail platforms.
 - Lots of work on improvements for simple-card, DaVinci and the Renesas
   rcar drivers.
 - New drivers for Analog Devices ADAU1977, TI PCM512x and parts of the
   CSR SiRF SoC.

10 years agoMerge tag 'ttm-fixes-3.14-2014-03-12' of git://people.freedesktop.org/~thomash/linux...
Dave Airlie [Thu, 13 Mar 2014 07:31:24 +0000 (17:31 +1000)] 
Merge tag 'ttm-fixes-3.14-2014-03-12' of git://people.freedesktop.org/~thomash/linux into drm-fixes

Second pull request of 2014-03-12. The first one was requested to be canceled.

Rob's fix for oops on invalidate_caches() and a fix for a
performance regression.

* tag 'ttm-fixes-3.14-2014-03-12' of git://people.freedesktop.org/~thomash/linux:
  drm/ttm: don't oops if no invalidate_caches()
  drm/ttm: Work around performance regression with VM_PFNMAP

10 years agoMerge branch 'drm-fixes-3.14' of git://people.freedesktop.org/~agd5f/linux into drm...
Dave Airlie [Thu, 13 Mar 2014 07:31:01 +0000 (17:31 +1000)] 
Merge branch 'drm-fixes-3.14' of git://people.freedesktop.org/~agd5f/linux into drm-fixes

A few more radeon fixes.

* 'drm-fixes-3.14' of git://people.freedesktop.org/~agd5f/linux:
  drm/radeon/cik: properly set compute ring status on disable
  drm/radeon/cik: stop the sdma engines in the enable() function
  drm/radeon/cik: properly set sdma ring status on disable
  drm/radeon: fix runpm disabling on non-PX harder

10 years agoMerge tag 'vmwgfx-fixes-3.14-2014-03-13' of git://people.freedesktop.org/~thomash...
Dave Airlie [Thu, 13 Mar 2014 07:30:31 +0000 (17:30 +1000)] 
Merge tag 'vmwgfx-fixes-3.14-2014-03-13' of git://people.freedesktop.org/~thomash/linux into drm-fixes

Pull request of 2014-03-13

one minor fix for new hw

* tag 'vmwgfx-fixes-3.14-2014-03-13' of git://people.freedesktop.org/~thomash/linux:
  drm/vmwgfx: Fix a surface reference corner-case in legacy emulation mode

10 years agodrm/vmwgfx: Fix a surface reference corner-case in legacy emulation mode
Thomas Hellstrom [Wed, 12 Mar 2014 19:42:31 +0000 (20:42 +0100)] 
drm/vmwgfx: Fix a surface reference corner-case in legacy emulation mode

If running on a gb-object capable device with a non-gb capable surface
exporter (X server) and a gb capable surface referencing client (GL driver),
the referencing client expects to find a shareable backing buffer attached to
the surface at reference time. This may not be the case if the surface has
not yet been validated. This would cause the surface reference IOCTL to
return an error.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jakob Bornecrantz <jakob@vmware.com>
10 years agoMerge tag 'pci-v3.14-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaa...
Linus Torvalds [Thu, 13 Mar 2014 01:29:34 +0000 (18:29 -0700)] 
Merge tag 'pci-v3.14-fixes-2' of git://git./linux/kernel/git/helgaas/pci

Pull PCI fixes from Bjorn Helgaas:
 "These are two important regression fixes for bugs we've introduced so
  far in v3.14.

  One of the resource allocation changes from the merge window is broken
  for 32-bit kernels where we don't use _CRS for PCI host bridges
  (mostly pre-2008 machines), so there's a fix for that.

  The INTx enable change we put in after the merge window turned out to
  break pciehp because we re-enable INTx on the hotplug bridge, which
  apparently breaks MSI for future hotplug events"

* tag 'pci-v3.14-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:
  PCI: Don't check resource_size() in pci_bus_alloc_resource()
  PCI: Enable INTx in pci_reenable_device() only when MSI/MSI-X not enabled

10 years agoMerge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Linus Torvalds [Thu, 13 Mar 2014 00:27:23 +0000 (17:27 -0700)] 
Merge tag 'for-linus' of git://git./virt/kvm/kvm

Pull KVM fixes from Paolo Bonzini:
 "The ARM patch fixes a build breakage with randconfig.  The x86 one
  fixes Windows guests on AMD processors"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
  KVM: SVM: fix cr8 intercept window
  ARM: KVM: fix non-VGIC compilation

10 years agoACPI / init: Invoke early ACPI initialization later
Rafael J. Wysocki [Wed, 12 Mar 2014 23:22:58 +0000 (00:22 +0100)] 
ACPI / init: Invoke early ACPI initialization later

Commit 73f7d1ca3263 (ACPI / init: Run acpi_early_init() before
timekeeping_init()) optimistically moved the early ACPI initialization
before timekeeping_init(), but that didn't work, because it broke fast
TSC calibration for Julian Wollrath on Thinkpad x121e (and most likely
for others too).  The reason is that acpi_early_init() enables the SCI
and that interferes with the fast TSC calibration mechanism.

Thus follow the original idea to execute acpi_early_init() before
efi_enter_virtual_mode() to help the EFI people for now and we can
revisit the other problem that commit 73f7d1ca3263 attempted to
address in the future (if really necessary).

Fixes: 73f7d1ca3263 (ACPI / init: Run acpi_early_init() before timekeeping_init())
Reported-by: Julian Wollrath <jwollrath@web.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agocpufreq: Skip current frequency initialization for ->setpolicy drivers
Rafael J. Wysocki [Wed, 12 Mar 2014 20:49:33 +0000 (21:49 +0100)] 
cpufreq: Skip current frequency initialization for ->setpolicy drivers

After commit da60ce9f2fac (cpufreq: call cpufreq_driver->get() after
calling ->init()) __cpufreq_add_dev() sometimes fails for CPUs handled
by intel_pstate, because that driver may return 0 from its ->get()
callback if it has not run long enough to collect enough samples on the
given CPU.  That didn't happen before commit da60ce9f2fac which added
policy->cur initialization to __cpufreq_add_dev() to help reduce code
duplication in other cpufreq drivers.

However, the code added by commit da60ce9f2fac need not be executed
for cpufreq drivers having the ->setpolicy callback defined, because
the subsequent invocation of cpufreq_set_policy() will use that
callback to initialize the policy anyway and it doesn't need
policy->cur to be initialized upfront.  The analogous code in
cpufreq_update_policy() is also unnecessary for cpufreq drivers
having ->setpolicy set and may be skipped for them as well.

Since intel_pstate provides ->setpolicy, skipping the upfront
policy->cur initialization for cpufreq drivers with that callback
set will cover intel_pstate and the problem it's been having after
commit da60ce9f2fac will be addressed.

Fixes: da60ce9f2fac (cpufreq: call cpufreq_driver->get() after calling ->init())
References: https://bugzilla.kernel.org/show_bug.cgi?id=71931
Reported-and-tested-by: Patrik Lundquist <patrik.lundquist@gmail.com>
Acked-by: Dirk Brandewie <dirk.j.brandewie@intel.com>
Cc: 3.13+ <stable@vger.kernel.org> # 3.13+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agoMerge remote-tracking branches 'asoc/topic/wm8996', 'asoc/topic/wm9081' and 'asoc...
Mark Brown [Wed, 12 Mar 2014 23:05:31 +0000 (23:05 +0000)] 
Merge remote-tracking branches 'asoc/topic/wm8996', 'asoc/topic/wm9081' and 'asoc/topic/wm9705' into asoc-next

10 years agoMerge remote-tracking branches 'asoc/topic/wm8961', 'asoc/topic/wm8962', 'asoc/topic...
Mark Brown [Wed, 12 Mar 2014 23:05:18 +0000 (23:05 +0000)] 
Merge remote-tracking branches 'asoc/topic/wm8961', 'asoc/topic/wm8962', 'asoc/topic/wm8974', 'asoc/topic/wm8983', 'asoc/topic/wm8988', 'asoc/topic/wm8990', 'asoc/topic/wm8991', 'asoc/topic/wm8993' and 'asoc/topic/wm8994' into asoc-next

10 years agoMerge remote-tracking branches 'asoc/topic/wm5100', 'asoc/topic/wm8523', 'asoc/topic...
Mark Brown [Wed, 12 Mar 2014 23:05:05 +0000 (23:05 +0000)] 
Merge remote-tracking branches 'asoc/topic/wm5100', 'asoc/topic/wm8523', 'asoc/topic/wm8731', 'asoc/topic/wm8737', 'asoc/topic/wm8741', 'asoc/topic/wm8753', 'asoc/topic/wm8903', 'asoc/topic/wm8904', 'asoc/topic/wm8940' and 'asoc/topic/wm8955' into asoc-next

10 years agoMerge remote-tracking branches 'asoc/topic/tlv320aic23', 'asoc/topic/tlv320aic26...
Mark Brown [Wed, 12 Mar 2014 23:04:52 +0000 (23:04 +0000)] 
Merge remote-tracking branches 'asoc/topic/tlv320aic23', 'asoc/topic/tlv320aic26', 'asoc/topic/tlv320aic32x4', 'asoc/topic/tlv320aic3x', 'asoc/topic/tlv320dac33', 'asoc/topic/twl4030', 'asoc/topic/twl6040', 'asoc/topic/uda1380', 'asoc/topic/wl1273' and 'asoc/topic/wm2200' into asoc-next

10 years agoMerge remote-tracking branches 'asoc/topic/samsung', 'asoc/topic/sgtl5000', 'asoc...
Mark Brown [Wed, 12 Mar 2014 23:04:42 +0000 (23:04 +0000)] 
Merge remote-tracking branches 'asoc/topic/samsung', 'asoc/topic/sgtl5000', 'asoc/topic/sh', 'asoc/topic/simple', 'asoc/topic/sirf', 'asoc/topic/sn95031', 'asoc/topic/ssm2602' and 'asoc/topic/stac9766' into asoc-next

10 years agoMerge remote-tracking branches 'asoc/topic/ml26124', 'asoc/topic/of', 'asoc/topic...
Mark Brown [Wed, 12 Mar 2014 23:04:35 +0000 (23:04 +0000)] 
Merge remote-tracking branches 'asoc/topic/ml26124', 'asoc/topic/of', 'asoc/topic/omap', 'asoc/topic/pxa' and 'asoc/topic/rcar' into asoc-next

10 years agoMerge remote-tracking branches 'asoc/topic/da7213', 'asoc/topic/da732x', 'asoc/topic...
Mark Brown [Wed, 12 Mar 2014 23:04:22 +0000 (23:04 +0000)] 
Merge remote-tracking branches 'asoc/topic/da7213', 'asoc/topic/da732x', 'asoc/topic/da9055', 'asoc/topic/davinci', 'asoc/topic/fsl', 'asoc/topic/fsl-esai', 'asoc/topic/fsl-sai' and 'asoc/topic/fsl-spdif' into asoc-next

10 years agoMerge remote-tracking branches 'asoc/topic/arizona', 'asoc/topic/atmel', 'asoc/topic...
Mark Brown [Wed, 12 Mar 2014 23:04:15 +0000 (23:04 +0000)] 
Merge remote-tracking branches 'asoc/topic/arizona', 'asoc/topic/atmel', 'asoc/topic/cirrus', 'asoc/topic/cs4271', 'asoc/topic/cs42l51', 'asoc/topic/cs42l52', 'asoc/topic/cs42l73' and 'asoc/topic/da7210' into asoc-next

10 years agoMerge remote-tracking branches 'asoc/topic/88pm860x', 'asoc/topic/ad1836', 'asoc...
Mark Brown [Wed, 12 Mar 2014 23:04:09 +0000 (23:04 +0000)] 
Merge remote-tracking branches 'asoc/topic/88pm860x', 'asoc/topic/ad1836', 'asoc/topic/ad193x', 'asoc/topic/adau1977', 'asoc/topic/ak4104', 'asoc/topic/ak4641', 'asoc/topic/ak4671' and 'asoc/topic/alc5623' into asoc-next

10 years agoMerge remote-tracking branch 'asoc/topic/pcm512x' into asoc-next
Mark Brown [Wed, 12 Mar 2014 23:04:08 +0000 (23:04 +0000)] 
Merge remote-tracking branch 'asoc/topic/pcm512x' into asoc-next

10 years agoMerge remote-tracking branch 'asoc/topic/pcm1792a' into asoc-next
Mark Brown [Wed, 12 Mar 2014 23:04:07 +0000 (23:04 +0000)] 
Merge remote-tracking branch 'asoc/topic/pcm1792a' into asoc-next

10 years agoMerge remote-tracking branch 'asoc/topic/pcm1681' into asoc-next
Mark Brown [Wed, 12 Mar 2014 23:04:07 +0000 (23:04 +0000)] 
Merge remote-tracking branch 'asoc/topic/pcm1681' into asoc-next

10 years agoMerge remote-tracking branch 'asoc/topic/enum' into asoc-next
Mark Brown [Wed, 12 Mar 2014 23:03:59 +0000 (23:03 +0000)] 
Merge remote-tracking branch 'asoc/topic/enum' into asoc-next

10 years agoMerge remote-tracking branch 'asoc/topic/dapm' into asoc-next
Mark Brown [Wed, 12 Mar 2014 23:03:58 +0000 (23:03 +0000)] 
Merge remote-tracking branch 'asoc/topic/dapm' into asoc-next

10 years agoMerge remote-tracking branch 'asoc/topic/core' into asoc-next
Mark Brown [Wed, 12 Mar 2014 23:03:58 +0000 (23:03 +0000)] 
Merge remote-tracking branch 'asoc/topic/core' into asoc-next

10 years agoMerge remote-tracking branch 'asoc/fix/spear' into asoc-linus
Mark Brown [Wed, 12 Mar 2014 23:03:57 +0000 (23:03 +0000)] 
Merge remote-tracking branch 'asoc/fix/spear' into asoc-linus

10 years agoMerge tag 'asoc-v3.14-rc6' into asoc-linus
Mark Brown [Wed, 12 Mar 2014 23:03:56 +0000 (23:03 +0000)] 
Merge tag 'asoc-v3.14-rc6' into asoc-linus

ASoC: Fixes for v3.14

A few things here:
 - Avoid memory leaks in error cases with DPCM, this code has never been
   that well tested in mainline due to the lack of mainline drivers but
   we now have one queued for the merge window!
 - Fix the N810 audio driver to load when booted with DT since the
   platform was converted to DT during the merge window.
 - Fixes for initialisation of some MFD drivers that are probably unused
   in mainline

# gpg: Signature made Mon 10 Mar 2014 18:21:32 GMT using RSA key ID 7EA229BD
# gpg: Good signature from "Mark Brown <broonie@sirena.org.uk>"
# gpg:                 aka "Mark Brown <broonie@debian.org>"
# gpg:                 aka "Mark Brown <broonie@kernel.org>"
# gpg:                 aka "Mark Brown <broonie@tardis.ed.ac.uk>"
# gpg:                 aka "Mark Brown <broonie@linaro.org>"
# gpg:                 aka "Mark Brown <Mark.Brown@linaro.org>"

10 years agoMerge tag 'sound-3.14-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai...
Linus Torvalds [Wed, 12 Mar 2014 22:36:04 +0000 (15:36 -0700)] 
Merge tag 'sound-3.14-rc7' of git://git./linux/kernel/git/tiwai/sound

Pull sound fixes from Takashi Iwai:
 "A few fixes for ASoC (N810 DT init fix, DPCM error path fix and a
  couple of MFD init fixes), and a fix for a Lenovo laptop.  All small
  and trivial fixes, suitable for rc7"

* tag 'sound-3.14-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
  ASoC: 88pm860: Fix IO setup
  ASoC: si476x: Fix IO setup
  ALSA: hda - Fix loud click noise with IdeaPad 410Y
  ASoC: pcm: free path list before exiting from error conditions
  ASoC: n810: fix init with DT boot

10 years agoASoC: tlv320aic23: Remove spurious bits per word setting
Mark Brown [Wed, 12 Mar 2014 13:32:34 +0000 (13:32 +0000)] 
ASoC: tlv320aic23: Remove spurious bits per word setting

regmap should handle any byte ordering issues required, it is looking for
a byte stream from the bus, so don't set 16 bits per word. This is likely
to have tested out OK due to use of an unmerged SPI controller driver.

Signed-off-by: Mark Brown <broonie@linaro.org>
Tested-by: Max Filippov <jcmvbkbc@gmail.com>
10 years agonet: socket: error on a negative msg_namelen
Matthew Leach [Tue, 11 Mar 2014 11:58:27 +0000 (11:58 +0000)] 
net: socket: error on a negative msg_namelen

When copying in a struct msghdr from the user, if the user has set the
msg_namelen parameter to a negative value it gets clamped to a valid
size due to a comparison between signed and unsigned values.

Ensure the syscall errors when the user passes in a negative value.

Signed-off-by: Matthew Leach <matthew.leach@arm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agodrm/radeon/cik: properly set compute ring status on disable
Alex Deucher [Wed, 12 Mar 2014 20:20:44 +0000 (16:20 -0400)] 
drm/radeon/cik: properly set compute ring status on disable

When we disable the rings, set the status properly.  If
not other code pathes may try and use the rings which are
not functional at this point.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
10 years agoMAINTAINERS: Add tools/net to NETWORKING [GENERAL]
Tobias Klauser [Tue, 11 Mar 2014 11:06:24 +0000 (12:06 +0100)] 
MAINTAINERS: Add tools/net to NETWORKING [GENERAL]

Make sure patches for these tools go to the netdev list as well.

References: https://marc.info/?l=linux-kernel&m=139450284501328&w=2
Cc: David S. Miller <davem@davemloft.net>
Cc: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agopacket: doc: Spelling s/than/that/
Geert Uytterhoeven [Tue, 11 Mar 2014 10:23:42 +0000 (11:23 +0100)] 
packet: doc: Spelling s/than/that/

Signed-off-by: Geert Uytterhoeven <geert+renesas@linux-m68k.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: netdev@vger.kernel.org
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agoMerge branch 'mlx4'
David S. Miller [Wed, 12 Mar 2014 20:13:17 +0000 (16:13 -0400)] 
Merge branch 'mlx4'

Or Gerlitz says:

====================
mlx4 fixes

These short series fixes two bugs related to the vxlan support and a
missing req module call for the IB driver which is needed to support
IB/RDMA over Ethernet.

Pathes done over the net tree, commit dd38743 "vlan: Set correct
source MAC address with TX VLAN offload enabled"
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agonet/mlx4_core: Load the IB driver when the device supports IBoE
Or Gerlitz [Wed, 12 Mar 2014 15:16:32 +0000 (17:16 +0200)] 
net/mlx4_core: Load the IB driver when the device supports IBoE

When checking what protocol drivers to load, the IB driver should be
requested also over Ethernet ports, if the device supports IBoE (RoCE).

Fixes: b046ffe 'net/mlx4_core: Load higher level modules according to ports type'
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agonet/mlx4_en: Handle vxlan steering rules for mac address changes
Or Gerlitz [Wed, 12 Mar 2014 15:16:31 +0000 (17:16 +0200)] 
net/mlx4_en: Handle vxlan steering rules for mac address changes

When the device mac address is changed, we must deregister the vxlan
steering rule associated with the previous mac, and register a new
steering rule using the new mac.

Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agonet/mlx4_core: Fix wrong dump of the vxlan offloads device capability
Or Gerlitz [Wed, 12 Mar 2014 15:16:30 +0000 (17:16 +0200)] 
net/mlx4_core: Fix wrong dump of the vxlan offloads device capability

Fix the value used to dump the vxlan offloads device capability to align
with the MLX4_DEV_CAP_FLAG2_yyy definition. While on that, add dump to
the IPoIB flow-steering device capability and fix small typo.

The vxlan cap value wasn't fully handled when a conflict was resolved
between MLX4_DEV_CAP_FLAG2_DMFS_IPOIB coming from the IB tree to
MLX4_DEV_CAP_FLAG2_VXLAN_OFFLOADS coming from net-next.

Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agodrm/radeon/cik: stop the sdma engines in the enable() function
Alex Deucher [Wed, 12 Mar 2014 19:26:34 +0000 (15:26 -0400)] 
drm/radeon/cik: stop the sdma engines in the enable() function

We always stop the rings when disabling the engines so just
call the stop functions directly from the sdma enable function.
This way the rings' status is set correctly on suspend so
there are no problems on resume.  Fixes resume failures that
result in acceleration getting disabled.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
10 years agodrm/radeon/cik: properly set sdma ring status on disable
Alex Deucher [Wed, 12 Mar 2014 19:15:58 +0000 (15:15 -0400)] 
drm/radeon/cik: properly set sdma ring status on disable

When we disable the rings, set the status properly.  If
not other code pathes may try and use the rings which are
not functional at this point.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
10 years agodrm/radeon: fix runpm disabling on non-PX harder
Alex Deucher [Tue, 11 Mar 2014 19:02:30 +0000 (15:02 -0400)] 
drm/radeon: fix runpm disabling on non-PX harder

Make sure runtime pm is disabled on non-PX hardware.
Should fix powerdown problems without displays attached.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
10 years agoxen-netback: use skb_is_gso in xenvif_start_xmit
Wei Liu [Tue, 11 Mar 2014 12:45:32 +0000 (12:45 +0000)] 
xen-netback: use skb_is_gso in xenvif_start_xmit

In 5bd076708 ("Xen-netback: Fix issue caused by using gso_type wrongly")
we use skb_is_gso to determine if we need an extra slot to accommodate
the SKB. There's similar error in interface.c. Change that to use
skb_is_gso as well.

Signed-off-by: Wei Liu <wei.liu2@citrix.com>
Cc: Annie Li <annie.li@oracle.com>
Cc: Ian Campbell <ian.campbell@citrix.com>
Cc: Paul Durrant <paul.durrant@citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agodrm/ttm: don't oops if no invalidate_caches()
Rob Clark [Wed, 12 Mar 2014 14:59:37 +0000 (10:59 -0400)] 
drm/ttm: don't oops if no invalidate_caches()

A few of the simpler TTM drivers (cirrus, ast, mgag200) do not implement
this function.  Yet can end up somehow with an evicted bo:

  BUG: unable to handle kernel NULL pointer dereference at           (null)
  IP: [<          (null)>]           (null)
  PGD 16e761067 PUD 16e6cf067 PMD 0
  Oops: 0010 [#1] SMP
  Modules linked in: bnep bluetooth rfkill fuse ip6t_rpfilter ip6t_REJECT ipt_REJECT xt_conntrack ebtable_nat ebtable_broute bridge stp llc ebtable_filter ebtables ip6table_nat nf_conntrack_ipv6 nf_defrag_ipv6 nf_nat_ipv6 ip6table_mangle ip6table_security ip6table_raw ip6table_filter ip6_tables iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack iptable_mangle iptable_security iptable_raw iptable_filter ip_tables sg btrfs zlib_deflate raid6_pq xor dm_queue_length iTCO_wdt iTCO_vendor_support coretemp kvm dcdbas dm_service_time microcode serio_raw pcspkr lpc_ich mfd_core i7core_edac edac_core ses enclosure ipmi_si ipmi_msghandler shpchp acpi_power_meter mperf nfsd auth_rpcgss nfs_acl lockd uinput sunrpc dm_multipath xfs libcrc32c ata_generic pata_acpi sr_mod cdrom
   sd_mod usb_storage mgag200 syscopyarea sysfillrect sysimgblt i2c_algo_bit lpfc drm_kms_helper ttm crc32c_intel ata_piix bfa drm ixgbe libata i2c_core mdio crc_t10dif ptp crct10dif_common pps_core scsi_transport_fc dca scsi_tgt megaraid_sas bnx2 dm_mirror dm_region_hash dm_log dm_mod
  CPU: 16 PID: 2572 Comm: X Not tainted 3.10.0-86.el7.x86_64 #1
  Hardware name: Dell Inc. PowerEdge R810/0H235N, BIOS 0.3.0 11/14/2009
  task: ffff8801799dabc0 ti: ffff88016c884000 task.ti: ffff88016c884000
  RIP: 0010:[<0000000000000000>]  [<          (null)>]           (null)
  RSP: 0018:ffff88016c885ad8  EFLAGS: 00010202
  RAX: ffffffffa04e94c0 RBX: ffff880178937a20 RCX: 0000000000000000
  RDX: 0000000000000000 RSI: 0000000000240004 RDI: ffff880178937a00
  RBP: ffff88016c885b60 R08: 00000000000171a0 R09: ffff88007cf171a0
  R10: ffffea0005842540 R11: ffffffff810487b9 R12: ffff880178937b30
  R13: ffff880178937a00 R14: ffff88016c885b78 R15: ffff880179929400
  FS:  00007f81ba2ef980(0000) GS:ffff88007cf00000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 0000000000000000 CR3: 000000016e763000 CR4: 00000000000007e0
  DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
  DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
  Stack:
   ffffffffa0306fae ffff8801799295c0 0000000000260004 0000000000000001
   ffff88016c885b60 ffffffffa0307669 00ff88007cf17738 ffff88017cf17700
   ffff880178937a00 ffff880100000000 ffff880100000000 0000000079929400
  Call Trace:
   [<ffffffffa0306fae>] ? ttm_bo_handle_move_mem+0x54e/0x5b0 [ttm]
   [<ffffffffa0307669>] ? ttm_bo_mem_space+0x169/0x340 [ttm]
   [<ffffffffa0307bd7>] ttm_bo_move_buffer+0x117/0x130 [ttm]
   [<ffffffff81130001>] ? perf_event_init_context+0x141/0x220
   [<ffffffffa0307cb1>] ttm_bo_validate+0xc1/0x130 [ttm]
   [<ffffffffa04e7377>] mgag200_bo_pin+0x87/0xc0 [mgag200]
   [<ffffffffa04e56c4>] mga_crtc_cursor_set+0x474/0xbb0 [mgag200]
   [<ffffffff811971d2>] ? __mem_cgroup_commit_charge+0x152/0x3b0
   [<ffffffff815c4182>] ? mutex_lock+0x12/0x2f
   [<ffffffffa0201433>] drm_mode_cursor_common+0x123/0x170 [drm]
   [<ffffffffa0205231>] drm_mode_cursor_ioctl+0x41/0x50 [drm]
   [<ffffffffa01f5ca2>] drm_ioctl+0x502/0x630 [drm]
   [<ffffffff815cbab4>] ? __do_page_fault+0x1f4/0x510
   [<ffffffff8101cb68>] ? __restore_xstate_sig+0x218/0x4f0
   [<ffffffff811b4445>] do_vfs_ioctl+0x2e5/0x4d0
   [<ffffffff8124488e>] ? file_has_perm+0x8e/0xa0
   [<ffffffff811b46b1>] SyS_ioctl+0x81/0xa0
   [<ffffffff815d05d9>] system_call_fastpath+0x16/0x1b
  Code:  Bad RIP value.
  RIP  [<          (null)>]           (null)
   RSP <ffff88016c885ad8>
  CR2: 0000000000000000

Signed-off-by: Rob Clark <rclark@redhat.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com>
Cc: stable@vger.kernel.org
10 years agodm cache: fix access beyond end of origin device
Heinz Mauelshagen [Wed, 12 Mar 2014 15:13:39 +0000 (16:13 +0100)] 
dm cache: fix access beyond end of origin device

In order to avoid wasting cache space a partial block at the end of the
origin device is not cached.  Unfortunately, the check for such a
partial block at the end of the origin device was flawed.

Fix accesses beyond the end of the origin device that occured due to
attempted promotion of an undetected partial block by:

- initializing the per bio data struct to allow cache_end_io to work properly
- recognizing access to the partial block at the end of the origin device
- avoiding out of bounds access to the discard bitset

Otherwise, users can experience errors like the following:

 attempt to access beyond end of device
 dm-5: rw=0, want=20971520, limit=20971456
 ...
 device-mapper: cache: promotion failed; couldn't copy block

Signed-off-by: Heinz Mauelshagen <heinzm@redhat.com>
Acked-by: Joe Thornber <ejt@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Cc: stable@vger.kernel.org
10 years agodm cache: fix truncation bug when copying a block to/from >2TB fast device
Heinz Mauelshagen [Tue, 11 Mar 2014 23:40:05 +0000 (00:40 +0100)] 
dm cache: fix truncation bug when copying a block to/from >2TB fast device

During demotion or promotion to a cache's >2TB fast device we must not
truncate the cache block's associated sector to 32bits.  The 32bit
temporary result of from_cblock() caused a 32bit multiplication when
calculating the sector of the fast device in issue_copy_real().

Use an intermediate 64bit type to store the 32bit from_cblock() to allow
for proper 64bit multiplication.

Here is an example of how this bug manifests on an ext4 filesystem:

 EXT4-fs error (device dm-0): ext4_mb_generate_buddy:756: group 17136, 32768 clusters in bitmap, 30688 in gd; block bitmap corrupt.
 JBD2: Spotted dirty metadata buffer (dev = dm-0, blocknr = 0). There's a risk of filesystem corruption in case of system crash.

Signed-off-by: Heinz Mauelshagen <heinzm@redhat.com>
Acked-by: Joe Thornber <ejt@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Cc: stable@vger.kernel.org
10 years agoKVM: SVM: fix cr8 intercept window
Radim Krčmář [Tue, 11 Mar 2014 18:11:18 +0000 (19:11 +0100)] 
KVM: SVM: fix cr8 intercept window

We always disable cr8 intercept in its handler, but only re-enable it
if handling KVM_REQ_EVENT, so there can be a window where we do not
intercept cr8 writes, which allows an interrupt to disrupt a higher
priority task.

Fix this by disabling intercepts in the same function that re-enables
them when needed. This fixes BSOD in Windows 2008.

Cc: <stable@vger.kernel.org>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Reviewed-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
10 years agoPCI: Don't check resource_size() in pci_bus_alloc_resource()
Bjorn Helgaas [Tue, 11 Mar 2014 20:23:12 +0000 (14:23 -0600)] 
PCI: Don't check resource_size() in pci_bus_alloc_resource()

Paul reported that after f75b99d5a77d ("PCI: Enforce bus address limits in
resource allocation") on a 32-bit kernel (CONFIG_PHYS_ADDR_T_64BIT not
set), intel-gtt complained "can't ioremap flush page - no chipset
flushing".  In addition, other PCI resource allocations, e.g., for bridge
windows, failed.

This happens because we incorrectly skip bus resources of
[mem 0x00000000-0xffffffff] because we think they are of size zero.
When resource_size_t is 32 bits wide, resource_size() on
[mem 0x00000000-0xffffffff] returns 0 because (r->end - r->start + 1)
overflows.

Therefore, we can't use "resource_size() == 0" to decide that allocation
from this resource will fail.  allocate_resource() should fail anyway if it
can't satisfy the address constraints, so we should just depend on that.

A [mem 0x00000000-0xffffffff] bus resource is obviously not really valid,
but we do fall back to it as a default when we don't have information about
host bridge apertures.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=71611
Fixes: f75b99d5a77d PCI: Enforce bus address limits in resource allocation
Reported-and-tested-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
10 years agoPCI: Enable INTx in pci_reenable_device() only when MSI/MSI-X not enabled
Bjorn Helgaas [Tue, 11 Mar 2014 20:22:19 +0000 (14:22 -0600)] 
PCI: Enable INTx in pci_reenable_device() only when MSI/MSI-X not enabled

Andreas reported that after 1f42db786b14 ("PCI: Enable INTx if BIOS left
them disabled"), pciehp surprise removal stopped working.

This happens because pci_reenable_device() on the hotplug bridge (used in
the pciehp_configure_device() path) clears the Interrupt Disable bit, which
apparently breaks the bridge's MSI hotplug event reporting.

Previously we cleared the Interrupt Disable bit in do_pci_enable_device(),
which is used by both pci_enable_device() and pci_reenable_device().  But
we use pci_reenable_device() after the driver may have enabled MSI or
MSI-X, and we *set* Interrupt Disable as part of enabling MSI/MSI-X.

This patch clears Interrupt Disable only when MSI/MSI-X has not been
enabled.

Fixes: 1f42db786b14 PCI: Enable INTx if BIOS left them disabled
Link: https://bugzilla.kernel.org/show_bug.cgi?id=71691
Reported-and-tested-by: Andreas Noever <andreas.noever@gmail.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
CC: stable@vger.kernel.org
CC: Sarah Sharp <sarah.a.sharp@linux.intel.com>
10 years agodrm/ttm: Work around performance regression with VM_PFNMAP
Thomas Hellstrom [Wed, 12 Mar 2014 09:41:32 +0000 (10:41 +0100)] 
drm/ttm: Work around performance regression with VM_PFNMAP

A performance regression was introduced in TTM in linux 3.13 when we started using
VM_PFNMAP for shared mappings. In theory this should've been faster due to
less page book-keeping but it appears like VM_PFNMAP + x86 PAT + write-combine
is a particularly cpu-hungry combination, as seen by largely increased
cpu-usage on r200 GL video playback.

Until we've sorted out why, revert to always use VM_MIXEDMAP.
Reference: freedesktop.org bugzilla bug #75719

Reported-and-tested-by: <smoki00790@gmail.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Cc: stable@vger.kernel.org
10 years agoASoC: Fix use after free
Lars-Peter Clausen [Wed, 12 Mar 2014 07:34:39 +0000 (08:34 +0100)] 
ASoC: Fix use after free

Freeing the current list element while iterating over the list will cause a use
after free since the iterator function will still use the current element to
look up the next. Use list_for_each_safe() and remove the element from the list
before freeing it to avoid this.

Fixes: 1438c2f60b ("ASoC: Add a per component dai list")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@linaro.org>
10 years ago[SCSI] storvsc: NULL pointer dereference fix
Ales Novak [Thu, 27 Feb 2014 10:03:30 +0000 (11:03 +0100)] 
[SCSI] storvsc: NULL pointer dereference fix

If the initialization of storvsc fails, the storvsc_device_destroy()
causes NULL pointer dereference.

storvsc_bus_scan()
  scsi_scan_target()
    __scsi_scan_target()
      scsi_probe_and_add_lun(hostdata=NULL)
        scsi_alloc_sdev(hostdata=NULL)

  sdev->hostdata = hostdata

  now the host allocation fails

          __scsi_remove_device(sdev)

  calls sdev->host->hostt->slave_destroy() ==
  storvsc_device_destroy(sdev)
    access of sdev->hostdata->request_mempool

Signed-off-by: Ales Novak <alnovak@suse.cz>
Signed-off-by: Thomas Abraham <tabraham@suse.com>
Reviewed-by: Jiri Kosina <jkosina@suse.cz>
Acked-by: K. Y. Srinivasan <kys@microsoft.com>
Cc: stable@vger.kernel.org
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
10 years agor8169: fix the incorrect tx descriptor version
hayeswang [Tue, 11 Mar 2014 07:11:59 +0000 (15:11 +0800)] 
r8169: fix the incorrect tx descriptor version

The tx descriptor version of RTL8111B belong to RTL_TD_0.

Signed-off-by: Hayes Wang <hayeswang@realtek.com>
Acked-by: Francois Romieu <romieu@fr.zoreil.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agotools/net/Makefile: Define PACKAGE to fix build problems
Markos Chandras [Tue, 11 Mar 2014 08:49:39 +0000 (08:49 +0000)] 
tools/net/Makefile: Define PACKAGE to fix build problems

Fixes the following build problem with binutils-2.24

gcc -Wall -O2   -c -o bpf_jit_disasm.o bpf_jit_disasm.c
In file included from bpf_jit_disasm.c:25:0:
/usr/include/bfd.h:35:2: error: #error config.h must be included
before this header
 #error config.h must be included before this header

This is similar to commit 3ce711a6abc27abce1554e1d671a8762b7187690
"perf tools: bfd.h/libbfd detection fails with recent binutils"

See: https://sourceware.org/bugzilla/show_bug.cgi?id=14243

CC: David S. Miller <davem@davemloft.net>
CC: Daniel Borkmann <dborkman@redhat.com>
CC: netdev@vger.kernel.org
Acked-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: Markos Chandras <markos.chandras@imgtec.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agox86: bpf_jit: support negative offsets
Alexei Starovoitov [Mon, 10 Mar 2014 22:56:51 +0000 (15:56 -0700)] 
x86: bpf_jit: support negative offsets

Commit a998d4342337 claimed to introduce negative offset support to x86 jit,
but it couldn't be working, since at the time of the execution
of LD+ABS or LD+IND instructions via call into
bpf_internal_load_pointer_neg_helper() the %edx (3rd argument of this func)
had junk value instead of access size in bytes (1 or 2 or 4).

Store size into %edx instead of %ecx (what original commit intended to do)

Fixes: a998d4342337 ("bpf jit: Let the x86 jit handle negative offsets")
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
Cc: Jan Seiffert <kaffeemonster@googlemail.com>
Cc: Eric Dumazet <edumazet@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agobridge: multicast: enable snooping on general queries only
Linus Lüssing [Mon, 10 Mar 2014 21:25:25 +0000 (22:25 +0100)] 
bridge: multicast: enable snooping on general queries only

Without this check someone could easily create a denial of service
by injecting multicast-specific queries to enable the bridge
snooping part if no real querier issuing periodic general queries
is present on the link which would result in the bridge wrongly
shutting down ports for multicast traffic as the bridge did not learn
about these listeners.

With this patch the snooping code is enabled upon receiving valid,
general queries only.

Signed-off-by: Linus Lüssing <linus.luessing@web.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agobridge: multicast: add sanity check for general query destination
Linus Lüssing [Mon, 10 Mar 2014 21:25:24 +0000 (22:25 +0100)] 
bridge: multicast: add sanity check for general query destination

General IGMP and MLD queries are supposed to have the multicast
link-local all-nodes address as their destination according to RFC2236
section 9, RFC3376 section 4.1.12/9.1, RFC2710 section 8 and RFC3810
section 5.1.15.

Without this check, such malformed IGMP/MLD queries can result in a
denial of service: The queries are ignored by most IGMP/MLD listeners
therefore they will not respond with an IGMP/MLD report. However,
without this patch these malformed MLD queries would enable the
snooping part in the bridge code, potentially shutting down the
according ports towards these hosts for multicast traffic as the
bridge did not learn about these listeners.

Reported-by: Jan Stancek <jstancek@redhat.com>
Signed-off-by: Linus Lüssing <linus.luessing@web.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agotcp: tcp_release_cb() should release socket ownership
Eric Dumazet [Mon, 10 Mar 2014 16:50:11 +0000 (09:50 -0700)] 
tcp: tcp_release_cb() should release socket ownership

Lars Persson reported following deadlock :

-000 |M:0x0:0x802B6AF8(asm) <-- arch_spin_lock
-001 |tcp_v4_rcv(skb = 0x8BD527A0) <-- sk = 0x8BE6B2A0
-002 |ip_local_deliver_finish(skb = 0x8BD527A0)
-003 |__netif_receive_skb_core(skb = 0x8BD527A0, ?)
-004 |netif_receive_skb(skb = 0x8BD527A0)
-005 |elk_poll(napi = 0x8C770500, budget = 64)
-006 |net_rx_action(?)
-007 |__do_softirq()
-008 |do_softirq()
-009 |local_bh_enable()
-010 |tcp_rcv_established(sk = 0x8BE6B2A0, skb = 0x87D3A9E0, th = 0x814EBE14, ?)
-011 |tcp_v4_do_rcv(sk = 0x8BE6B2A0, skb = 0x87D3A9E0)
-012 |tcp_delack_timer_handler(sk = 0x8BE6B2A0)
-013 |tcp_release_cb(sk = 0x8BE6B2A0)
-014 |release_sock(sk = 0x8BE6B2A0)
-015 |tcp_sendmsg(?, sk = 0x8BE6B2A0, ?, ?)
-016 |sock_sendmsg(sock = 0x8518C4C0, msg = 0x87D8DAA8, size = 4096)
-017 |kernel_sendmsg(?, ?, ?, ?, size = 4096)
-018 |smb_send_kvec()
-019 |smb_send_rqst(server = 0x87C4D400, rqst = 0x87D8DBA0)
-020 |cifs_call_async()
-021 |cifs_async_writev(wdata = 0x87FD6580)
-022 |cifs_writepages(mapping = 0x852096E4, wbc = 0x87D8DC88)
-023 |__writeback_single_inode(inode = 0x852095D0, wbc = 0x87D8DC88)
-024 |writeback_sb_inodes(sb = 0x87D6D800, wb = 0x87E4A9C0, work = 0x87D8DD88)
-025 |__writeback_inodes_wb(wb = 0x87E4A9C0, work = 0x87D8DD88)
-026 |wb_writeback(wb = 0x87E4A9C0, work = 0x87D8DD88)
-027 |wb_do_writeback(wb = 0x87E4A9C0, force_wait = 0)
-028 |bdi_writeback_workfn(work = 0x87E4A9CC)
-029 |process_one_work(worker = 0x8B045880, work = 0x87E4A9CC)
-030 |worker_thread(__worker = 0x8B045880)
-031 |kthread(_create = 0x87CADD90)
-032 |ret_from_kernel_thread(asm)

Bug occurs because __tcp_checksum_complete_user() enables BH, assuming
it is running from softirq context.

Lars trace involved a NIC without RX checksum support but other points
are problematic as well, like the prequeue stuff.

Problem is triggered by a timer, that found socket being owned by user.

tcp_release_cb() should call tcp_write_timer_handler() or
tcp_delack_timer_handler() in the appropriate context :

BH disabled and socket lock held, but 'owned' field cleared,
as if they were running from timer handlers.

Fixes: 6f458dfb4092 ("tcp: improve latencies of timer triggered events")
Reported-by: Lars Persson <lars.persson@axis.com>
Tested-by: Lars Persson <lars.persson@axis.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agoMerge branch 'skb_frags'
David S. Miller [Tue, 11 Mar 2014 20:26:46 +0000 (16:26 -0400)] 
Merge branch 'skb_frags'

Michael S. Tsirkin says:

====================
skbuff: fix skb_segment with zero copy skbs

This fixes a bug in skb_segment where it moves frags
between skbs without orphaning them.
This causes userspace to assume it's safe to
reuse the buffer, and receiver gets corrupted data.
This further might leak information from the
transmitter on the wire.

To fix track which skb does a copied frag belong
to, and orphan frags when copying them.

As we are tracking multiple skbs here, using
short names (skb,nskb,fskb,skb_frag,frag) becomes confusing.
So before adding another one, I refactor these names
slightly.

Patch is split out to make it easier to
verify that all trasformations are trivially correct.

The problem was observed in the field,
so I think that the patch is necessary on stable
as well.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
10 years agoskbuff: skb_segment: orphan frags before copying
Michael S. Tsirkin [Mon, 10 Mar 2014 17:28:08 +0000 (19:28 +0200)] 
skbuff: skb_segment: orphan frags before copying

skb_segment copies frags around, so we need
to copy them carefully to avoid accessing
user memory after reporting completion to userspace
through a callback.

skb_segment doesn't normally happen on datapath:
TSO needs to be disabled - so disabling zero copy
in this case does not look like a big deal.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agoskbuff: skb_segment: s/fskb/list_skb/
Michael S. Tsirkin [Mon, 10 Mar 2014 17:27:59 +0000 (19:27 +0200)] 
skbuff: skb_segment: s/fskb/list_skb/

fskb is unrelated to frag: it's coming from
frag_list. Rename it list_skb to avoid confusion.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agoskbuff: skb_segment: s/skb/head_skb/
Michael S. Tsirkin [Mon, 10 Mar 2014 16:29:19 +0000 (18:29 +0200)] 
skbuff: skb_segment: s/skb/head_skb/

rename local variable to make it easier to tell at a glance that we are
dealing with a head skb.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agoskbuff: skb_segment: s/skb_frag/frag/
Michael S. Tsirkin [Mon, 10 Mar 2014 16:29:14 +0000 (18:29 +0200)] 
skbuff: skb_segment: s/skb_frag/frag/

skb_frag can in fact point at either skb
or fskb so rename it generally "frag".

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agoskbuff: skb_segment: s/frag/nskb_frag/
Michael S. Tsirkin [Mon, 10 Mar 2014 16:29:04 +0000 (18:29 +0200)] 
skbuff: skb_segment: s/frag/nskb_frag/

frag points at nskb, so name it appropriately

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agoPNP / ACPI: proper handling of ACPI IO/Memory resource parsing failures
Zhang Rui [Tue, 11 Mar 2014 14:40:27 +0000 (22:40 +0800)] 
PNP / ACPI: proper handling of ACPI IO/Memory resource parsing failures

Before commit b355cee88e3b (ACPI / resources: ignore invalid ACPI
device resources), if acpi_dev_resource_memory()/acpi_dev_resource_io()
returns false, it means the the resource is not a memeory/IO resource.

But after commit b355cee88e3b, those functions return false if the
given memory/IO resource entry is invalid (the length of the resource
is zero).

This breaks pnpacpi_allocated_resource(), because it now recognizes
the invalid memory/io resources as resources of unknown type.  Thus
users see confusing warning messages on machines with zero length
ACPI memory/IO resources.

Fix the problem by rearranging pnpacpi_allocated_resource() so that
it calls acpi_dev_resource_memory() for memory type and IO type
resources only, respectively.

Fixes: b355cee88e3b (ACPI / resources: ignore invalid ACPI device resources)
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Reported-and-tested-by: Markus Trippelsdorf <markus@trippelsdorf.de>
Reported-and-tested-by: Julian Wollrath <jwollrath@web.de>
Reported-and-tested-by: Paul Bolle <pebolle@tiscali.nl>
[rjw: Changelog]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
10 years agoMerge branch 'stmmac'
David S. Miller [Tue, 11 Mar 2014 20:15:29 +0000 (16:15 -0400)] 
Merge branch 'stmmac'

Giuseppe Cavallaro says:

====================
stmmac fixes: EEE and chained mode

These patches are to fix some new problems in the STMMAC driver.

Mandatory changes are for EEE that needs to be disabled if not supported
and for the chain mode that is broken and the kernel panics if this mode
is enabled.

v3: removed a patch from my previous set that touched the stmmac_tx path
    that has not to be applied. Other patches for cleaning-up will be
    sent on top of net-next git repo.

v4: do not surround the defaul buffer selection using Koption and adopt
    a default to 1536bytes.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agostmmac: dwmac-sti: fix broken STiD127 compatibility
Giuseppe CAVALLARO [Mon, 10 Mar 2014 12:40:34 +0000 (13:40 +0100)] 
stmmac: dwmac-sti: fix broken STiD127 compatibility

This is to fix the compatibility to the STiD127 SoC.

Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Cc: Srinivas Kandagatla <srinivas.kandagatla@st.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agostmmac: fix chained mode
Giuseppe CAVALLARO [Mon, 10 Mar 2014 12:40:33 +0000 (13:40 +0100)] 
stmmac: fix chained mode

This patch is to fix the chain mode that was broken
and generated a panic. This patch reviews the chain/ring
modes now shaing the same structure and taking care
about the pointers and callbacks.

Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agostmmac: fix and better tune the default buffer sizes
Giuseppe CAVALLARO [Mon, 10 Mar 2014 12:40:32 +0000 (13:40 +0100)] 
stmmac: fix and better tune the default buffer sizes

This patch is to fix and tune the default buffer sizes.
It reduces the default bufsize used by the driver from
4KiB to 1536 bytes.

Patch has been tested on both ARM and SH4 platform based.

Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agostmmac: disable at run-time the EEE if not supported
Giuseppe CAVALLARO [Mon, 10 Mar 2014 12:40:31 +0000 (13:40 +0100)] 
stmmac: disable at run-time the EEE if not supported

This patch is to disable the EEE (so HW and timers)
for example when the phy communicates that the EEE
can be supported anymore.

Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agovmxnet3: fix netpoll race condition
Neil Horman [Mon, 10 Mar 2014 10:55:55 +0000 (06:55 -0400)] 
vmxnet3: fix netpoll race condition

vmxnet3's netpoll driver is incorrectly coded.  It directly calls
vmxnet3_do_poll, which is the driver internal napi poll routine.  As the netpoll
controller method doesn't block real napi polls in any way, there is a potential
for race conditions in which the netpoll controller method and the napi poll
method run concurrently.  The result is data corruption causing panics such as this
one recently observed:
PID: 1371   TASK: ffff88023762caa0  CPU: 1   COMMAND: "rs:main Q:Reg"
 #0 [ffff88023abd5780] machine_kexec at ffffffff81038f3b
 #1 [ffff88023abd57e0] crash_kexec at ffffffff810c5d92
 #2 [ffff88023abd58b0] oops_end at ffffffff8152b570
 #3 [ffff88023abd58e0] die at ffffffff81010e0b
 #4 [ffff88023abd5910] do_trap at ffffffff8152add4
 #5 [ffff88023abd5970] do_invalid_op at ffffffff8100cf95
 #6 [ffff88023abd5a10] invalid_op at ffffffff8100bf9b
    [exception RIP: vmxnet3_rq_rx_complete+1968]
    RIP: ffffffffa00f1e80  RSP: ffff88023abd5ac8  RFLAGS: 00010086
    RAX: 0000000000000000  RBX: ffff88023b5dcee0  RCX: 00000000000000c0
    RDX: 0000000000000000  RSI: 00000000000005f2  RDI: ffff88023b5dcee0
    RBP: ffff88023abd5b48   R8: 0000000000000000   R9: ffff88023a3b6048
    R10: 0000000000000000  R11: 0000000000000002  R12: ffff8802398d4cd8
    R13: ffff88023af35140  R14: ffff88023b60c890  R15: 0000000000000000
    ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
 #7 [ffff88023abd5b50] vmxnet3_do_poll at ffffffffa00f204a [vmxnet3]
 #8 [ffff88023abd5b80] vmxnet3_netpoll at ffffffffa00f209c [vmxnet3]
 #9 [ffff88023abd5ba0] netpoll_poll_dev at ffffffff81472bb7

The fix is to do as other drivers do, and have the poll controller call the top
half interrupt handler, which schedules a napi poll properly to recieve frames

Tested by myself, successfully.

Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
CC: Shreyas Bhatewara <sbhatewara@vmware.com>
CC: "VMware, Inc." <pv-drivers@vmware.com>
CC: "David S. Miller" <davem@davemloft.net>
CC: stable@vger.kernel.org
Reviewed-by: Shreyas N Bhatewara <sbhatewara@vmware.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agox86, fpu: Check tsk_used_math() in kernel_fpu_end() for eager FPU
Suresh Siddha [Mon, 3 Feb 2014 06:56:23 +0000 (22:56 -0800)] 
x86, fpu: Check tsk_used_math() in kernel_fpu_end() for eager FPU

For non-eager fpu mode, thread's fpu state is allocated during the first
fpu usage (in the context of device not available exception). This
(math_state_restore()) can be a blocking call and hence we enable
interrupts (which were originally disabled when the exception happened),
allocate memory and disable interrupts etc.

But the eager-fpu mode, call's the same math_state_restore() from
kernel_fpu_end(). The assumption being that tsk_used_math() is always
set for the eager-fpu mode and thus avoid the code path of enabling
interrupts, allocating fpu state using blocking call and disable
interrupts etc.

But the below issue was noticed by Maarten Baert, Nate Eldredge and
few others:

If a user process dumps core on an ecrypt fs while aesni-intel is loaded,
we get a BUG() in __find_get_block() complaining that it was called with
interrupts disabled; then all further accesses to our ecrypt fs hang
and we have to reboot.

The aesni-intel code (encrypting the core file that we are writing) needs
the FPU and quite properly wraps its code in kernel_fpu_{begin,end}(),
the latter of which calls math_state_restore(). So after kernel_fpu_end(),
interrupts may be disabled, which nobody seems to expect, and they stay
that way until we eventually get to __find_get_block() which barfs.

For eager fpu, most the time, tsk_used_math() is true. At few instances
during thread exit, signal return handling etc, tsk_used_math() might
be false.

In kernel_fpu_end(), for eager-fpu, call math_state_restore()
only if tsk_used_math() is set. Otherwise, don't bother. Kernel code
path which cleared tsk_used_math() knows what needs to be done
with the fpu state.

Reported-by: Maarten Baert <maarten-baert@hotmail.com>
Reported-by: Nate Eldredge <nate@thatsmathematics.com>
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Suresh Siddha <sbsiddha@gmail.com>
Link: http://lkml.kernel.org/r/1391410583.3801.6.camel@europa
Cc: George Spelvin <linux@horizon.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
10 years agoMerge branch 'for-next' of git://git.samba.org/sfrench/cifs-2.6
Linus Torvalds [Tue, 11 Mar 2014 18:53:42 +0000 (11:53 -0700)] 
Merge branch 'for-next' of git://git.samba.org/sfrench/cifs-2.6

Pull CIFS fixes from Steve French:
 "A fix for the problem which Al spotted in cifs_writev and a followup
  (noticed when fixing CVE-2014-0069) patch to ensure that cifs never
  sends more than the smb frame length over the socket (as we saw with
  that cifs_iovec_write problem that Jeff fixed last month)"

* 'for-next' of git://git.samba.org/sfrench/cifs-2.6:
  cifs: mask off top byte in get_rfc1002_length()
  cifs: sanity check length of data to send before sending
  CIFS: Fix wrong pos argument of cifs_find_lock_conflict

10 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm...
Linus Torvalds [Tue, 11 Mar 2014 17:17:50 +0000 (10:17 -0700)] 
Merge branch 'for-linus' of git://git./linux/kernel/git/ebiederm/user-namespace

Pull audit namespace fixes from Eric Biederman:
 "Starting with 3.14-rc1 the audit code is faulty (think oopses and
  races) with respect to how it computes the network namespace of which
  socket to reply to, and I happened to notice by chance when reading
  through the code.

  My testing and the automated build bots don't find any problems with
  these fixes"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace:
  audit: Update kdoc for audit_send_reply and audit_list_rules_send
  audit: Send replies in the proper network namespace.
  audit: Use struct net not pid_t to remember the network namespce to reply in

10 years agox86: Remove CONFIG_X86_OOSTORE
Dave Jones [Mon, 10 Mar 2014 23:32:22 +0000 (19:32 -0400)] 
x86: Remove CONFIG_X86_OOSTORE

This was an optimization that made memcpy type benchmarks a little
faster on ancient (Circa 1998) IDT Winchip CPUs.  In real-life
workloads, it wasn't even noticable, and I doubt anyone is running
benchmarks on 16 year old silicon any more.

Given this code has likely seen very little use over the last decade,
let's just remove it.

Signed-off-by: Dave Jones <davej@fedoraproject.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
10 years agoALSA: sis7019: Simplify dependencies
Jean Delvare [Tue, 11 Mar 2014 13:30:07 +0000 (14:30 +0100)] 
ALSA: sis7019: Simplify dependencies

Kconfig symbol X86_32 was introduced in October 2005, it's about time
to use it :-)

Signed-off-by: Jean Delvare <jdelvare@suse.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
10 years agoALSA: cs5535audio: Also needed on MIPS
Jean Delvare [Tue, 11 Mar 2014 10:38:34 +0000 (11:38 +0100)] 
ALSA: cs5535audio: Also needed on MIPS

The CS5536 companion chipset is not only used on 32-bit x86 systems as
I originally thought, it is also used on MIPS Loongson/Lemote 2
systems. So let the snd-cs5535audio driver be built on MIPS too.

Signed-off-by: Jean Delvare <jdelvare@suse.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
10 years agoperf/x86: Fix leak in uncore_type_init failure paths
Dave Jones [Thu, 6 Mar 2014 17:20:28 +0000 (12:20 -0500)] 
perf/x86: Fix leak in uncore_type_init failure paths

The error path of uncore_type_init() frees up any allocations
that were made along the way, but it relies upon type->pmus
being set, which only happens if the function succeeds. As
type->pmus remains null in this case, the call to
uncore_type_exit will do nothing.

Moving the assignment earlier will allow us to actually free
those allocations should something go awry.

Signed-off-by: Dave Jones <davej@fedoraproject.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20140306172028.GA552@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
10 years agosched/clock: Prevent tracing recursion in sched_clock_cpu()
Fernando Luis Vazquez Cao [Thu, 6 Mar 2014 05:25:28 +0000 (14:25 +0900)] 
sched/clock: Prevent tracing recursion in sched_clock_cpu()

Prevent tracing of preempt_disable/enable() in sched_clock_cpu().
When CONFIG_DEBUG_PREEMPT is enabled, preempt_disable/enable() are
traced and this causes trace_clock() users (and probably others) to
go into an infinite recursion. Systems with a stable sched_clock()
are not affected.

This problem is similar to that fixed by upstream commit 95ef1e52922
("KVM guest: prevent tracing recursion with kvmclock").

Signed-off-by: Fernando Luis Vazquez Cao <fernando@oss.ntt.co.jp>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/1394083528.4524.3.camel@nexus
Signed-off-by: Ingo Molnar <mingo@kernel.org>
10 years agostop_machine: Fix^2 race between stop_two_cpus() and stop_cpus()
Peter Zijlstra [Fri, 28 Feb 2014 12:39:05 +0000 (13:39 +0100)] 
stop_machine: Fix^2 race between stop_two_cpus() and stop_cpus()

We must use smp_call_function_single(.wait=1) for the
irq_cpu_stop_queue_work() to ensure the queueing is actually done under
stop_cpus_lock. Without this we could have dropped the lock by the time
we do the queueing and get the race we tried to fix.

Fixes: 7053ea1a34fa ("stop_machine: Fix race between stop_two_cpus() and stop_cpus()")
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: Prarit Bhargava <prarit@redhat.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Link: http://lkml.kernel.org/r/20140228123905.GK3104@twins.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
10 years agosched/deadline: Deny unprivileged users to set/change SCHED_DEADLINE policy
Juri Lelli [Mon, 3 Mar 2014 11:09:21 +0000 (12:09 +0100)] 
sched/deadline: Deny unprivileged users to set/change SCHED_DEADLINE policy

Deny the use of SCHED_DEADLINE policy to unprivileged users.
Even if root users can set the policy for normal users, we
don't want the latter to be able to change their parameters
(safest behavior).

Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1393844961-18097-1-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
10 years agoASoC: core: Fix check before setting default I/O up try regmap
Xiubo Li [Tue, 11 Mar 2014 04:43:22 +0000 (12:43 +0800)] 
ASoC: core: Fix check before setting default I/O up try regmap

Since the CODEC driver could specify its own I/O(read and write)
while registering the CODEC for some reason, maybe the MFDs is
used, etc.

So just do check it, if they are not specified by CODEC driver
then try to set up the default regmap I/O if regmap is used.

Signed-off-by: Xiubo Li <Li.Xiubo@freescale.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
10 years agoASoC: io: New signature for snd_soc_codec_set_cache_io()
Xiubo Li [Tue, 11 Mar 2014 04:43:21 +0000 (12:43 +0800)] 
ASoC: io: New signature for snd_soc_codec_set_cache_io()

Now that all users have been converted to regmap and the config.reg_bits
and config.val_bits can be setted by each user through regmap core API.
So these two params are redundant here.

Since the only control type that left is SND_SOC_REGMAP, so remove it. Drop
the control params and add struct regmap *regmap to simplify the code.

Signed-off-by: Xiubo Li <Li.Xiubo@freescale.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
10 years agoASoC: codec: Simplify ASoC probe code.
Xiubo Li [Tue, 11 Mar 2014 04:43:20 +0000 (12:43 +0800)] 
ASoC: codec: Simplify ASoC probe code.

For some CODEC drivers like who act as the MFDs children are ignored
by this patch.

Signed-off-by: Xiubo Li <Li.Xiubo@freescale.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
10 years agoMerge branches 'topic/ad193x', 'topic/tlv320aic23', 'topic/tlv320aic32x4', 'topic...
Mark Brown [Tue, 11 Mar 2014 09:58:55 +0000 (09:58 +0000)] 
Merge branches 'topic/ad193x', 'topic/tlv320aic23', 'topic/tlv320aic32x4', 'topic/wm8991', 'fix/si476x' and 'fix/88pm860' of git://git./linux/kernel/git/broonie/sound into asoc-io

10 years agoMerge tag 'perf-urgent-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git...
Ingo Molnar [Tue, 11 Mar 2014 09:45:32 +0000 (10:45 +0100)] 
Merge tag 'perf-urgent-for-mingo' of git://git./linux/kernel/git/acme/linux into perf/urgent

Pull perf/urgent fixes from Arnaldo Carvalho de Melo:

 * Fix build of 'trace' in some systems due to using some architecture-specific
   signal numbers (Ben Hutchings)

 * Stop resolving when finding a map in in ip__resolve_ams, this way at least
   the DSO will be resolved when a symbol isn't (Don Zickus)

 * Fix crash in elf_section_by_name when not checking if some section string index
   is valid (Jiri Olsa)

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
10 years agoALSA: hda - initialize audio InfoFrame to be all zero
Mengdong Lin [Tue, 11 Mar 2014 21:12:52 +0000 (17:12 -0400)] 
ALSA: hda - initialize audio InfoFrame to be all zero

This patch initialized the local audio InfoFrame variable 'ai' to be all zero,
thus the data bytes will indicate "Refer to Stream Header" by default.

Signed-off-by: Mengdong Lin <mengdong.lin@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
10 years agoRevert "ALSA: emu10k1: Fix warning: "CCR" redefined"
Geert Uytterhoeven [Mon, 10 Mar 2014 20:20:55 +0000 (21:20 +0100)] 
Revert "ALSA: emu10k1: Fix warning: "CCR" redefined"

This reverts commit 83fc3bc09518d42e8f5073e2a65884701dfadf19.

sh-specific "CCR" and "CCR2" have been prefixed by "SH_" in commit
a5f6ea29f9a918403629f8369ae55fac6b09cb53 ('sh: prefix sh-specific "CCR" and
"CCR2" by "SH_"').

Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
10 years agoALSA: cs553*: Fix dependencies
Jean Delvare [Mon, 10 Mar 2014 20:12:26 +0000 (21:12 +0100)] 
ALSA: cs553*: Fix dependencies

The CS5530, CS5535 and CS5536 chipsets are companions of the Geode
series of processors, which are 32-bit x86 processors. So the
snd-cs5530 and snd-cs5535audio drivers are only needed on this
architecture, except for build testing purpose.

Signed-off-by: Jean Delvare <jdelvare@suse.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
10 years agoMerge tag 'asoc-v3.14-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie...
Takashi Iwai [Tue, 11 Mar 2014 06:50:33 +0000 (07:50 +0100)] 
Merge tag 'asoc-v3.14-rc6' of git://git./linux/kernel/git/broonie/sound into for-linus

ASoC: Fixes for v3.14

A few things here:
 - Avoid memory leaks in error cases with DPCM, this code has never been
   that well tested in mainline due to the lack of mainline drivers but
   we now have one queued for the merge window!
 - Fix the N810 audio driver to load when booted with DT since the
   platform was converted to DT during the merge window.
 - Fixes for initialisation of some MFD drivers that are probably unused
   in mainline

10 years agovlan: Set correct source MAC address with TX VLAN offload enabled
Peter Boström [Mon, 10 Mar 2014 15:17:15 +0000 (16:17 +0100)] 
vlan: Set correct source MAC address with TX VLAN offload enabled

With TX VLAN offload enabled the source MAC address for frames sent using the
VLAN interface is currently set to the address of the real interface. This is
wrong since the VLAN interface may be configured with a different address.

The bug was introduced in commit 2205369a314e12fcec4781cc73ac9c08fc2b47de
("vlan: Fix header ops passthru when doing TX VLAN offload.").

This patch sets the source address before calling the create function of the
real interface.

Signed-off-by: Peter Boström <peter.bostrom@netrounds.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agoXen-netback: Fix issue caused by using gso_type wrongly
Annie Li [Mon, 10 Mar 2014 14:58:34 +0000 (22:58 +0800)] 
Xen-netback: Fix issue caused by using gso_type wrongly

Current netback uses gso_type to check whether the skb contains
gso offload, and this is wrong. Gso_size is the right one to
check gso existence, and gso_type is only used to check gso type.

Some skbs contains nonzero gso_type and zero gso_size, current
netback would treat these skbs as gso and create wrong response
for this. This also causes ssh failure to domu from other server.

V2: use skb_is_gso function as Paul Durrant suggested

Signed-off-by: Annie Li <annie.li@oracle.com>
Acked-by: Wei Liu <wei.liu2@citrix.com>
Reviewed-by: Paul Durrant <paul.durrant@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
10 years agoMerge branch 'akpm' (patches from Andrew Morton)
Linus Torvalds [Tue, 11 Mar 2014 00:26:36 +0000 (17:26 -0700)] 
Merge branch 'akpm' (patches from Andrew Morton)

Merge misc fixes from Andrew Morton:
 "Nine fixes"

* emailed patches from Andrew Morton akpm@linux-foundation.org>:
  cris: convert ffs from an object-like macro to a function-like macro
  hfsplus: add HFSX subfolder count support
  tools/testing/selftests/ipc/msgque.c: handle msgget failure return correctly
  MAINTAINERS: blackfin: add git repository
  revert "kallsyms: fix absolute addresses for kASLR"
  mm/Kconfig: fix URL for zsmalloc benchmark
  fs/proc/base.c: fix GPF in /proc/$PID/map_files
  mm/compaction: break out of loop on !PageBuddy in isolate_freepages_block
  mm: fix GFP_THISNODE callers and clarify

10 years agocris: convert ffs from an object-like macro to a function-like macro
Geert Uytterhoeven [Mon, 10 Mar 2014 22:49:52 +0000 (15:49 -0700)] 
cris: convert ffs from an object-like macro to a function-like macro

This avoids bad interactions with code using identifiers called "ffs":

  drivers/usb/gadget/f_fs.c: In function 'ffsmod_init':
  drivers/usb/gadget/f_fs.c:2693:494: error: 'ffsusb_func' undeclared (first use in this function)
  drivers/usb/gadget/f_fs.c:2693:494: note: each undeclared identifier is reported only once for each function it appears in
  drivers/usb/gadget/f_fs.c: In function 'ffsmod_exit':
  drivers/usb/gadget/f_fs.c:2693:677: error: 'ffsusb_func' undeclared (first use in this function)
  drivers/usb/gadget/f_fs.c: At top level:
  drivers/usb/gadget/f_fs.c:2693:35: warning: 'kernel_ffsusb_func' defined but not used [-Wunused-variable]
  drivers/usb/gadget/f_fs.c: In function 'ffsmod_init':
  drivers/usb/gadget/f_fs.c:2693:15: warning: control reaches end of non-void function [-Wreturn-type]

See http://kisskb.ellerman.id.au/kisskb/buildresult/10715817/

Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Mikael Starvik <starvik@axis.com>
Cc: Jesper Nilsson <jesper.nilsson@axis.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
10 years agohfsplus: add HFSX subfolder count support
Sergei Antonov [Mon, 10 Mar 2014 22:49:51 +0000 (15:49 -0700)] 
hfsplus: add HFSX subfolder count support

Adds support for HFSX 'HasFolderCount' flag and a corresponding
'folderCount' field in folder records.  (For reference see
HFS_FOLDERCOUNT and kHFSHasFolderCountBit/kHFSHasFolderCountMask in
Apple's source code.)

Ignoring subfolder count leads to fs errors found by Mac:

  ...
  Checking catalog hierarchy.
  HasFolderCount flag needs to be set (id = 105)
  (It should be 0x10 instead of 0)
  Incorrect folder count in a directory (id = 2)
  (It should be 7 instead of 6)
  ...

Steps to reproduce:
 Format with "newfs_hfs -s /dev/diskXXX".
 Mount in Linux.
 Create a new directory in root.
 Unmount.
 Run "fsck_hfs /dev/diskXXX".

The patch handles directory creation, deletion, and rename.

Signed-off-by: Sergei Antonov <saproj@gmail.com>
Reviewed-by: Vyacheslav Dubeyko <slava@dubeyko.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
10 years agotools/testing/selftests/ipc/msgque.c: handle msgget failure return correctly
Colin Ian King [Mon, 10 Mar 2014 22:49:50 +0000 (15:49 -0700)] 
tools/testing/selftests/ipc/msgque.c: handle msgget failure return correctly

A failed msgget causes the test to return an uninitialised value in ret.
Assign ret to -errno on error exit.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Davidlohr Bueso <davidlohr@hp.com>
Cc: Stanislav Kinsbursky <skinsbursky@parallels.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
10 years agoMAINTAINERS: blackfin: add git repository
Michael Opdenacker [Mon, 10 Mar 2014 22:49:49 +0000 (15:49 -0700)] 
MAINTAINERS: blackfin: add git repository

Add the git repository currently in use for blackfin architecture
development.

This information was obtained from Steven Miao.

Signed-off-by: Michael Opdenacker <michael.opdenacker@free-electrons.com>
Cc: Steven Miao <realmz6@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This page took 0.055685 seconds and 5 git commands to generate.